I buy electric stove and found that thermostat placement doesn’t make any sense. enter image description here The thermostat used seem like one in iron, in that case, the iron heater connected directly to thermostat via ceramic ring. In this electric stove seem like it regulate air temperature inside. At first I think this product is bad design but I found many product install thermostat the same way (at the case without direct contact to heater). At this point I wonder do I think the wrong way or whole product that sell in the market got wrong design.

  • 1
    \$\begingroup\$ You can easily test this using a thermometer at the heater location, itself. And yes, this looks like a terrible position except that given enough time the air will equilibrate. But it means that if you start cooking early enough, you will very likely overheat the food and that could ruin the dish. For this device, perhaps best to let it sit for a while before using it to cook anything. You can also test the idea using an oil that is sensitive to temperature -- extra virgin olive oil comes to mind, which will color if the temperature is excessive. Probably a cost vs market expectation thing. \$\endgroup\$
    – jonk
    May 1, 2022 at 18:19
  • 5
    \$\begingroup\$ That's a "simmerstat", not a thermostat. It regulates the duty-cycle of the power applied to the heating element, not the temperature. \$\endgroup\$
    – brhans
    May 1, 2022 at 18:40
  • 3
    \$\begingroup\$ have you thought about a kitchen stove? ... the controls are not near the element either \$\endgroup\$
    – jsotola
    May 1, 2022 at 20:53

5 Answers 5


The thermostat used ...

It's not a thermostat. It's a power controller.

... seem like one in iron,

Correct. Clothes irons use the same type of power controller.

In this electric stove seem like it regulate air temperature inside.

enter image description here

Image source.

No. The contacts are initially closed applying heat to the load. A bimetallic strip then starts to bend and toggles open after a time switching off the load. It then cools and reconnects and the cycle starts again. The duty-cycle (and hence the average power) is adjusted by the knob which determines how far the bimetallic strip has to bend before toggling.

At this point I wonder do I think the wrong way or whole product that sell in the market got wrong design.

No, you seem to think that you cook in pots based on temperature. You don't. You cook based on power. For example, what temperature would you set the ring at to boil water? 100°C, 105°C, 125°C? The water in the pot will never get above 100°C. Instead you determine the rate of boiling (simmer, etc.) by controlling the power.

... and found that thermostat placement doesn’t make any sense.

It should be clear now that your initial premise was incorrect and that the design does make sense. Controlling the power is how gas cookers work too. There are no thermostats on the hob.

Note that an oven will use a proper thermostat - usually a bulb and capillary tube. The liquid in the tube expands and trips the switch at a temperature set by the rotation of the thermostat control knob.

  • \$\begingroup\$ Found something similar to a datasheet here (they have more than 100 models...): robertshaw.com/WorkArea/DownloadAsset.aspx?id=2147485598. There is a nice angle vs duty cycle graph (from Excel, it seems...). \$\endgroup\$
    – devnull
    May 1, 2022 at 22:30
  • \$\begingroup\$ Normally the ambient load power is far greater than than the losses in the switch by several orders of magnitude with proximity to heater makes this a load sensitive heater thermostat contrary to your assumption. \$\endgroup\$ May 1, 2022 at 23:08
  • \$\begingroup\$ Note that your link defines the part description contradicts your assertion that it is not a thermostat \$\endgroup\$ May 1, 2022 at 23:12

It isn't really acting as a thermostat at all. It's an electro-mechanical oscillator, where the proportion of the time spent on and off is adjustable.

It has a bimetallic switch, like a true thermostat. But the heat to drive it comes from the current passing through the switch, not the stove element.

  • 2
    \$\begingroup\$ @TonyStewartEE75 It probably isn't constant with age. There's no calibration on the controls, just numbers. That one looks like it goes from OFF to 5. Similar controls are used on cookers with four rings, and the control may be nowhere near the ring that it's controlling. \$\endgroup\$
    – Simon B
    May 1, 2022 at 22:24
  • 2
    \$\begingroup\$ Mine is better. It goes to 6. \$\endgroup\$
    – Transistor
    May 1, 2022 at 22:25
  • 2
    \$\begingroup\$ en.wikipedia.org/wiki/Infinite_switch \$\endgroup\$
    – Drew
    May 2, 2022 at 3:11
  • 1
    \$\begingroup\$ @TonyStewartEE75 Luckily, there is a feedback that compensate for those variations: the person using the stove to cook, who will adjust the knob based on the effects on the food. :) \$\endgroup\$
    – Redy000
    May 2, 2022 at 10:10
  • 1
    \$\begingroup\$ @TonyStewartEE75 How can you assume the heat from the switch is constant with age? You cannot. But feel free to open any of the tens-hundreds of millions of electromechanically controlled cooktops in the US alone, and you'll find exactly this approach. It works well enough. There's a human in the loop: if the pot is about to boil over, the cook turns the heat down. That's why they tell people not to leave their stove unattended. The whole thing depends on there being a human in the loop, by design. \$\endgroup\$ May 2, 2022 at 11:40

The temperature of an electric iron is controlled by a bimetallic strip which, on getting bent by the heat from the sole plate, de-energises the heating element by opening a pair of contacts.

The temperature at which that happens is determined by the pressure of the regulating knob on the bimetallic strip.

As the sole plate cools, the contacts close and the process repeats.

The regulating knob dial has positions marked e.g. 'Rayon ', 'Silk', 'Wool', 'Cotton' and 'Linen' with the highest temperature for linen and the lowest for rayon.

enter image description here

On the other hand the electric stove is controlled by an 'energy regulator' or 'simmerstat'.

It comprises a bimetallic strip that is heated by a coil wound over it and not by the stove heating element.

When the bimetallic strip bends, a pair of contacts open and de-energise the bimetal heating coil and also the stove heating element.

The point at which the contact opens is determined by the pressure of the regulating knob on the bimetallic strip.

As the bimetallic strip cools, the contacts close and the process repeats.

In effect, the energy regulator carries out heating element 'duty-cycle control'.

The regulating knob dial has graduations marked 0 - 100 % duty, with the contacts remaining continuously on at 100% and continuously off at 0 %.

enter image description here


This is a thermostat with sufficiently remote sensing of the conducted and radiated heat so that it does not over-react to the heater element rapid temperature swings.

Your assumptions have no merit without measurements on a stovetop with and without a thermal load.

Power contact switches have a limited lifespan for rated and actual current. The goal ought to be to have a MTBF of 10 years with normal operational use. Thus the cycling rate must be slow enough to not prematurely wear out. I assume the thermal heatsink mass and space provides the necessary thermal steady state temperature ripple of some x degrees of load temp.

The goal ought to be to sense the temperature of the load and not have too close a feedback from the heater , so location and thermal coupling or resistance must be designed to compromise sensing the heater and the load.

In this case the contacts appear to be the heater, not the sensor and the control is a variable spring loaded bi-metallic switch oscillator with thermal stove-edge feedback. SO I would expect the duty cycle of roughly a minute cycle or so to change with the load and changes in steady state heat sink and source of power.

The choice of bi-metals are quite large and the trigger temperature and average power may be influenced by nearby heat and ambient temperature of the mass between the target and the heater. The design is quite complex in reality for the thermal properties of the switch. So thermal properties of the switch, convection and conduction each have some influence in controlling the desire behaviour to have a more uniform result. enter image description here

When making any assumptions about a design, one must apply Thermodynamics, Joules Law and analyze the conduction and convection thermal resistance of the system. The switch is not isolated nor uniform in this application and the duty cycle of switching must not be too slow nor too fast.

  • 1
    \$\begingroup\$ This is a rather circuitous way to describe a circuit element in use so widespread that there are about a billion of them in active use in the US, and that's a conservative estimate... Everyone can easily see the billions of electromechanical on-off switches in use, but those simmerstats are in the same league. An extremely common part in spite of its relative bulk. These things are just an order or two of magnitude shy of the CPU/MCU count in domestic use. \$\endgroup\$ May 2, 2022 at 11:46
  • 1
    \$\begingroup\$ TY @Kubahasn'tforgottenMonica There certainly was alot more to the design than the eye can see. emsclad.com/fileadmin/Data/Divisions/EMS/Header/… Some are uniform thermally activated while most are resistively self-heated but biased by nearby ambient \$\endgroup\$ May 2, 2022 at 13:37
  • \$\begingroup\$ This is called a thermostatic switch and convection, conduction and self-heating are all part of the operation in many cases. But there may be a difference in the US "TEMPATURE" and the ROTW TEMPERature controls. ;) \$\endgroup\$ May 2, 2022 at 14:01

I think this kind of control depends on the temperature of the metal strips supporting the contacts, rather than sensing the surrounding air temperature, or the actual heating element temperature.


Not the answer you're looking for? Browse other questions tagged or ask your own question.